journal of dentistry 39 (2011) 341–350

available at www.sciencedirect.com

journal homepage: www.intl.elsevierhealth.com/journals/jden

Review

Acupuncture for treating temporomandibular joint disorders:

A systematic review and meta-analysis of randomized,
sham-controlled trials

Aram Jung a, Byung-Cheul Shin a,*, Myeong Soo Lee b, Hoseob Sim a, Edzard Ernst c
a
School of Korean Medicine, Pusan National University, Yangsan, South Korea
b
Brain Disease Research Centre, Korea Institute of Oriental Medicine, Daejeon, South Korea
c
Complementary Medicine, Peninsula Medical School, Universities of Exeter, Exeter, UK

article info abstract

Article history: Objective: The aim of this article was to assess the clinical evidence for or against acupunc-
Received 31 October 2010 ture and acupuncture-like therapies as treatments for temporomandibular joint disorder
Received in revised form (TMD).
13 February 2011 Data: This systematic review includes randomized clinical trials (RCTs) of acupuncture as a
Accepted 21 February 2011 treatment for TMD compared to sham acupuncture. The search terms were selected
according to medical subject heading (MeSH).
Sources: Systematic searches were conducted in 13 electronic databases up to July 2010;
Keywords: Medline, PubMed, The Cochrane Library 2010 (Issue 7), CINAHL, EMBASE, seven Korean
Acupuncture Medical Databases and a Chinese Medical Database.
Temporomandibular joint disorder Study selection: All parallel or cross-over RCTs of acupuncture for TMD were searched
Pain without language restrictions. Studies in which no clinical data and complex interventions
Systematic review were excluded. Finally, total of 7 RCTs met our inclusion criteria.
Meta-analysis Conclusions: In conclusion, our systematic review and meta-analysis demonstrate that the
evidence for acupuncture as a symptomatic treatment of TMD is limited. Further rigorous
studies are, however, required to establish beyond doubt whether acupuncture has thera-
peutic value for this indication.
# 2011 Elsevier Ltd. All rights reserved.

women than men.4,5 Its aetiology regards as multi-factorial,

1. Introduction
A temporomandibular joint disorder (TMD) is clinically
characterized by pain and dysfunction in the masticatory
muscles or temporomandibular joint (TMJ).1 TMDs are most
common form of chronic orofacial pain.2 The prevalence of
TMD in the U.S.A. is between 40% and 75%.2,3 TMD is common
in adults aged 20–50 years, and it is more prevalent amongst
structure-related, and controversial.2
Current medical interventions for the management of TMD
consist of jaw-appliance therapy, medications, physiotherapy,
home self-care and surgery.1 Noninvasive, nonsurgical thera-
pies are most commonly used for TMD.6 Though there is
controversy exists in reporting whether they were successful
for TMD treatments to date,6 however, clinical research and
experience have been reported TMD is successfully managed

* Corresponding author at: Division of Clinical Medicine, School of Korean Medicine, Pusan National University, Yangsan 626-870,
Kyungnam, South Korea. Tel.: +82 51 510 8482; fax: +82 51 510 8420.
E-mail addresses: drshinbc@gmail.com, drshinbc@pusan.ac.kr (B.-C. Shin).
0300-5712/$ – see front matter # 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.jdent.2011.02.006
342 journal of dentistry 39 (2011) 341–350
by one or another of current modalities once the correct
diagnosis and proper management are provided.7
A survey showed that 74% of TMD patients used comple-
mentary and alternative medicine (CAM) therapies. Most of
the respondents reported being most satisfied with the ‘‘hands
on’’ CAM therapies such as massage or acupuncture.8
Acupuncture has been claimed to be effective in TMD
treatments in the mechanism of pain reduction, anti-
inflammation, and neurohormonal effects.9,10
In acupuncture clinical research, it is difficult to draw a
reliable conclusion due to lack of appropriate placebo.11 Of
various placebo controls, ‘‘sham’’ procedure can be defined as
one performed on a control group to ensure that they have the
same experiences like real group subjects do.11 Furthermore,
several recent studies reported sham acupuncture interven-
tion is more relevance method than other physical placebos
due to the effect of excluded psychological effect.12,13 In one
neurologic study based on brain-imaging analysis, the
psychological responses to placebo analgesia were similar
to those that were elicited by the administration of ‘‘real’’
analgesic substances.14 Therefore, a sham control method is
more rigorous than other controls for identifying the specific
effects of acupuncture especially in pain.
Up to recently, four systematic reviews of acupuncture for
TMD are available.15–18 However, one review was out of date16
and secondly published one reviewed 4 sham controlled
randomized clinical trials (RCTs) without comprehensive
meta-analysis.17 The third one reviewed 19 RCTs (including
5 sham controlled) without meta-analysis15 and the last
published one did 9 RCTs (including 4 sham controlled) with
meta-analysis.18 All did not conduct full searches because
three of them restricted their searches only in English.16–18
Consequently, all were missing the important sham controlled
RCTs.19–21 In addition, none adopted subgroup analysis and
sensitivity analysis of meta-analysis which were strictly
recommended by PRISMA guideline.22 Thus, their conclusions
are unsafe, even can be biased.
Therefore, the aims of our review were to critically evaluate
the totality of the most rigorous clinical evidence for or against
the effectiveness of acupuncture-type treatments compared
to relevant sham one in patients with TMD and to provide a
comprehensive PRISMA-compliant systematic review with
sensitivity and subgroup analysis.
2. Methods
2.1. Data sources
The following electronic databases were searched from their
inceptions up to July 2010: Medline, PubMed, The Cochrane
Library 2010 (Issue 7), CINAHL, EMBASE, seven Korean Medical
Databases (DBPIA, OASIS, Korea Institute of Science and
Technology Information, National Assembly Library, Korean
Studies Information, The Journal of Korean Acupuncture and
Moxibustion Society, The Korean Journal of Meridian and
Acupoint) and a Chinese Medical Database (China Academic
Journal, www.cnki.co.kr).
The search phrases used were [temporomandibular or
temporomandibular joint or Temporomandibular joint disor-
der or TMD or Jaw disease or Craniomandibular disease or
{(myofascial pain) or (myofascial pain dysfunction) and TMD}]
and [acupuncture or Acup* or moxibustion or acupressure or
‘‘laser acupuncture’’ or ‘‘auricular acupuncture’’] in English,
Chinese and Korean. We selected these terms according to
medical subject heading (MeSH). In addition, the references in
all located articles were manually searched for further
relevant articles.
2.2. Study selection
2.2.1. Types of studies
This review included parallel or cross-over RCTs that assessed
the efficacy of acupuncture regardless of blinding, language
and type of reporting. Studies in which no clinical data were
reported were excluded. We also excluded complex interven-
tions in which acupuncture was not a sole treatment.
Dissertations and abstracts were included, if provided they
contained sufficient detail.
2.2.2. Types of patients
This study included subjects with TMD that was diagnosed by
any defined or specified diagnosis criteria, regardless of their
age, race and gender. The conditions of patients in the
included RCTs were classified into articular-, muscular or the
two types combined of TMD. Studies in which patients were
reported as having pain or functional symptoms in the jaw
muscle, temples, face, pre-auricular area, or in the ear were
included. Our study excluded patients with TMD found to be
caused by psychogenic, neurologic and metabolic disorders.
2.2.3. Types of interventions
Acupuncture is defined as the stimulation of acupuncture
points or trigger points by needles that pierce the skin, or by
heating the mugwort herb (moxibustion) in combination with
needles, or by electrical stimulation.23 Methods of stimulating
acupuncture points that do not involve needle insertion (e.g.,
laser, acupressure, moxibustion) are also included in this
review. We included only sham acupuncture (such as
penetrating or non-penetrating sham needle, non-activated
laser acupuncture) as controls.
2.2.4. Outcome measures
One of the following outcome measures was required for
inclusion: pain intensity or pain relief in TMJ measured by
visual analogue scale (VAS), verbal scale, or algometer. Other
clinically important outcomes included measured maximum
inter-incisal mouth opening (MO), or range of motion (ROM) of
TMJ, or response rate (responder vs. non-responder).
2.2.5. Data extraction and quality assessment
Copies of all articles of RCTs were obtained and read in full. All
articles were read by two independent reviewers (ARJ, BCS)
and data from the articles were validated and extracted
according to pre-defined criteria.24 Quality assessment was
assessed using the Cochrane risk of bias criteria. The quality of
the RCTs was addressed as follows amongst 6 domains: (1)
Was the allocation sequence adequately generated? (2) Was
allocation adequately concealed? (3) Was knowledge of the
allocated intervention adequately prevented during the study?
 journal of dentistry 39 (2011) 341–350 343

(4) Were incomplete outcome data adequately addressed? (5)
Are reports of the study free of suggestion of selective outcome
reporting? (6) Was the study apparently free of other problems
that could put it at a high risk of bias? As practitioner blinding
may not be possible due to the nature of acupuncture, patient
blinding was assumed for cases in which the control
intervention was indistinguishable from acupuncture, even
if the word ‘blinding’ did not occur in the report. The point for
evaluator blinding was only given if specified in the text. We
resolved any disagreements by referring to the trial report and
through discussion between reviewers (ARJ, BCS) and if
needed, by seeking the opinion of a third reviewer (MSL).
calculated using Cochrane Collaboration software [Review
Manager 5 (RevMan) Version 5.0.25 for Windows. Copenhagen:
The Nordic Cochrane Centre]. The variance of the change was
imputed using a correlation factor of 0.5 as suggested by the
Cochrane Collaboration.25 We then pooled data across studies
using random effect models if excessive statistical heteroge-
neity did not exist. The chi-square test and the Higgins I2 test
were used to assess heterogeneity.26 For more comprehensive
understanding of meta-analysis or if any kinds of heteroge-
neity exist, we conducted subgroup analysis or sensitivity
analysis additionally. Publication bias was assessed by funnel-
plot using the Cochrane software.

2.3. Data synthesis 2.4. Statistical analysis

To summarize the effects of acupuncture on each outcome, we
abstracted the risk estimates (relative risk: RR) for dichoto-
mous data and standardized mean difference (SMD) for
continuous data with 95% confidence interval (CI). RR or
[()TD$FIG]SMD or weight mean difference (WMD) with 95% CI were
To estimate sample size, we set the mean with standard
deviation from the difference of VAS between real and sham
groups. Sample size was estimated by 80% power (0.2 in beta
error) and 0.05 alpha error by using SigmaPlot software for
Windows Version 11.0 (Systat Softwere Inc., San Jose, CA, U.S.A.).

Publications identified (n=491)

Publications excluded after screening the abstract

and title (n =398). Reasons:
Not related to acupuncture (n=52)
Not related to TMD (n=111)
Not clinical studies (n=117)
Animal studies (n=11)
Case series (n=38)
Case reports (n=69)

Articles further evaluated by full text (n =93)

Excluded (n=34)
UCTs (n=18)
NRCTs (n=16)

RCT but excluded, because of (n=52)

Mixed treatments (n=18)
Not TMD patients (n=1)
Duplication (n=5)
Compared with another type of acupuncture
(n=4)
Compared with other active therapies
(n=23)
(6 trials originated from 2 trials, finally
n=19, summarized in Supplement 1)
Sham controlled but excluded because of
insufficient data (n=1, also reported in
Supplement 1)

7 RCTs met inclusion criteria

(needle acupuncture n=6, laser acupuncture
n=1)

Fig. 1 – Flow chart of the trial selection process. TMD: temporomandibular joint disorder; UCT: uncontrolled clinical trial;
NRCT: non-randomized controlled clinical trial, and RCT: randomized clinical trial.
 344
Table 1 – Summary of RCTs of acupuncture versus Sham acupuncture for TMD.
First author (year) Design/blinded Intervention Control group (regime) Main outcomes Intergroup AT method / Adverse event /
Randomized group (regime) differences acupoint/deqi risk of bias
Duration of TMD (week)
Types of TMD
Goddard (2002)29 2 parallel/DB (A) AT (30 min, (B) Sham AT (30 min once, (1) 100-mm VAS (1) P = 0.27, Fixed point/ n.r./Y,U,Y,Y,Y,U
18 once, n = 10) nonacupoint, penetrating (pain intensity) MD, 0.53 LI4, ST6/n.r.
>12 skin, 2–4 mm depth, n = 8) [ 1.48, 0.42]
Muscular

Smith (2007)31 2 parallel/DB (A) AT (20 min, (B) Sham AT (Park sham (1) 10-cm VAS (1) (A) MD;33.2 Fixed point/ST7/ n.r./Y,Y,Y,Y,Y,N
27 6 times for 3 needle, acupoint, non- (pain intensity) (B) MD;0.8a considered
weeks, n = 15) penetrating skin, 20 min (pre-post)
once, 6 times for 3

journal of dentistry 39 (2011) 341–350

weeks, n = 12)
>24 (2) MO (mm) (2) (A) MD; 3.2
(B) MD; 1.5a
(pre-post)
Combined (3) Muscle (3) P = 0.12,
tenderness RR, 4.801,35

Shen (2007)20 2 parallel/DB (A) AT (15 min, (B) Sham AT (1 cm distal to (1) 10-cm VAS (1) P = 0.06, Fixed point/ n.r./U,U,Y,Y,Y,Y
15 once, n = 9) LI4, non-penetrating, 15 min (pain intensity) MD, 1.09 LI4/n.r.
once, n = 6) [ 2.21, 0.04]
>12 (2) NRS (2) P = 0.22,
Muscular (facial pain) MD, 1.22
[ 3.16, 0.72]

Shen (2009)28 2 parallel/DB (A) AT (15 min, (B) Sham AT (1 cm distal to (1) 10-cm VAS (1) P = 0.24, Fixed point/ n.r./Y,U,Y,Y,Y,Y
28 once, n = 16) LI4, non-penetrating, 15 min (pain intensity) MD, 0.45 LI4/n.r.
once, n = 12) [ 1.21, 0.31]
n.r. (2) NRS (2) P = 0.84,
Muscular (facial pain) MD, 0.20
[ 2.14, 1.74]

Schmid-Schwap 2 parallel/DB (A) AT (20 min, (B) Sham laser (acupoint, (1) 100-mm VAS (1) P = 0.04, Fixed point/ no/U,Y,Y,Y,Y,Y
(2006)30 23 6 times for non-activated sham laser, (pain intensity) MD, 0.92 LI4, SI2,
3 weeks, n = 11) 20 min once, n = 12) [ 1.79, 0.05] SI3/n.r.
n.r. (2) MO (mm) (2) P = 0.10,
MD, 0.71
[ 0.14, 1.55]
Combined (3) Muscle (3) P = 0.002,
tenderness MD, 1.52
[ 2.47, 0.57]
 journal of dentistry 39 (2011) 341–350 345

RR = relative risk; MD: mean difference; RCT: randomized clinical trial; n.r.: not report; Risk of bias: yes (Y): low risk of bias; No (N): high risk of bias; Unclear (U); DB: double blind, this means patient
AT: acupuncture; MO: mouth opening; TMD: temporomandibular joint disorder; VAS: visual analogue scale (100 mm scale); NRS: numeric rating scale (0–10 point, 11point scale); NS: non-significant;
n.r./U,U,Y,Y,N,U
n.r./Y,U,Y,Y,Y,Y

3. Results

3.1. Study description

The searches identified 491 potentially relevant studies, of

which 484 were excluded (Fig. 1). Of the excluded studies, 24
RCTs were excluded because they included other active
ST6, SI18, SI3,

treatments as controls instead of sham one, or because of

Fixed point/
SI2, SI3/n.r.
point/LI4,

scanty data.27 Those studies are separately summarized

LI4/n.r.

without explanation in Supplement 1. Finally, 7 RCTs met

Fixed

our inclusion criteria. The key data are summarized in Table 1.

Three trials originated from U.S.A.,20,28,29 two from
Austria,21,30 one from U.K.,31 and one from Germany.19
[ 1.18, 0.47]

[ 1.55, 0.14]

A total of 141 patients were included in these studies (mean

[0.13, 44.41]
(1) P = 0.40,

(2) P = 0.10,

(1) P = 0.52,

(2) P = 0.56,
[0.40, 1.58]
MD, 0.35

MD, 0.70

sample size: 20, female:male = 19.1:1, mean age: 37.3 years).

RR, 0.80

RR, 2.40

The diagnostic methods of TMD used were two trials29,31

diagnosed the subjects by Research diagnosed criteria (RDC/
TMD),2 the others did not report which diagnostic systems
were used.19–21,28,30 In all studies, for the types of TMD, five
(1) 100-mm VAS

(1) 100-mm VAS

(pain intensity)

(pain intensity)

(2) Verbal scale

studies were muscular type,19–21,28,29 none was articular type,

(pain free %)

and two were combined type of both.30,31 Six trials evaluated

tenderness
(2) Muscle

needle acupuncture,20,21,28–31 and one trial evaluated laser

acupuncture.19 The duration of TMD was generally more than
12 weeks. The duration of total treatment ranged from 1
treatment to 3 weeks (mean: 1.4 weeks). The subjective
times/week, 6 sessions, n = 3)

outcome measures of these trials were the VAS, numeric

non-activated sham laser,

non-activated sham laser,

1session = 15 min each, 2
(B) Sham laser (acupoint,

(B) Sham laser (acupoint,

rating scale (NRS), muscle tenderness (4 point scale) or verbal

scale (response rate of pain free) and an objective one was MO
(millimetre). The baseline comparison of TMD symptoms was
n.r., once, n = 12)

reported in five trials,20,21,28–30 but not in two.19,31

Outcomes cannot be estimated because this study does not provide standard deviation.

3.2. Description of acupuncture treatment

Of the 7 total studies, six studies comparatively tested needle

acupuncture against penetrating sham acupuncture,29 non-
penetrating sham acupuncture,20,28,31 or sham laser acupunc-
2 treatment/week,

ture,21,30 whilst the remaining study tested laser acupuncture

3 weeks, n = 4)
6 sessions for
15 min each,
once, n = 11)

(A) Laser AT

against sham laser acupuncture.19

(A) AT (n.r.,

A total of 6 acupuncture points were used (see Fig. 2). Of

them, LI 4 was used the most (34 times/total 91 points),
followed by SI 3 (19/91), ST 6 (13/91), SI 18 (12/91), SI 2 (7/91), and
ST 7 (6/91). Three of these 6 adopted acupoints were located in
the face, especially in the affected mandible (50%), whilst three
acupoints were in the hand (50%).

3.3. Adverse event

2 parallel/DB

2 parallel/DB

Only one RCT reported adverse event data of acupuncture and

Muscular

Muscular

sham acupuncture and this stated that there were no adverse

events.30 None of others mentioned adverse events.
n.r.

n.r.
23

3.4. Risk of bias in the included studies

and assessor-blind.
Katsoulis (2010)19

The methodological quality of the RCTs was variable. Four

Simma (2009)21

studies described an appropriate method of sequence genera-

tion.21,28,29,31 One used a random number table,29 and the others
used a computerized randomization method.21,28,31 The other
three studies did not clearly report how the allocation sequence
a

was generated. Allocation was concealed in two studies.30,31 All

346 journal of dentistry 39 (2011) 341–350
[()TD$FIG]
Fig. 2 – Acupuncture points used in included studies.
RCTs adopted both patient- and assessor-blinding and analysed
all participants without dropouts and withdrawals. Only two
studies reported details about the dropouts and withdra-
wals.20,31 Five studies were considered quality in high (low risk
of bias)21,28–31 and two studies reported blinding but unclear
randomization methods without appropriate allocation con-
cealment thus the quality of them was in moderate.19,20 In
addition, there was no publication bias because of symmetrical
distribution by funnel-plot (data was not shown).
3.5. Outcomes
3.5.1. VAS for pain intensity
All seven RCTs reported the results of manual acupuncture
contrasted with sham acupuncture in terms of pain intensity by
VAS.19–21,28–31 The point of time for measuring the VAS was
referring to ‘now’ in six trials,20,21,28–31 that of the other one to
‘over last 14 days’.19 VAS scale on subjective pain in patients was
evaluated in same region in all studies, where were the main
area surrounding the TMJ, generally face, jaw and masseter
muscle.19–21,28–31 The time for evaluating VAS scale was
immediately after treatment in 6 studies,20,21,28–31 the other
one evaluated it after 16 weeks.19 Five trials showed favourable
effects of acupuncture, whilst the others did not. The pooled
meta-analysis of data showed significant improvements in pain
intensity for VAS (5 studies, n = 107, WMD, 13.63; 95% CI of
21.16 to 6.10, P = 0.0004, heterogeneity: x2 = 1.46, P = 0.83,
I = 0%).20,21,28–30 A subgroup analysis of VAS for pain intensity by
2
conditions showed favourable effects on real acupuncture
superior to sham acupuncture regardless of types of TMD;
muscular type (4 studies, n = 84, WMD, 14.28; 95% CI of 24.66
to 3.90, P = 0.007, heterogeneity: x2 = 1.42, P = 0.70, I2 = 0%), or
combined type of TMD (1 study, n = 23, WMD, 12.90; 95% CI of
23.83 to 1.97, P = 0.02) (Fig. 3A1). By the number of treatments,
included four RCTs that used only one-time acupuncture
treatment also showed favourable effects of acupuncture in
pain intensity by VAS20,21,28,29 (4 studies, n = 84, WMD, 14.28;
95% CI of 24.66 to 3.90, P = 0.007, heterogeneity: x2 = 1.42,
P = 0.70, I2 = 0%) (Fig. 3A2). Further subgroup analysis showed
acupuncture to be superior to non-penetrating sham control
methods in pain reduction by VAS in 4 studies (n = 89, WMD,
13.73; 95% CI of 21.78 to 5.67, P = 0.0008, heterogeneity:
x2 = 1.45, P = 0.69, I2 = 0%).20,21,28,30 No significant difference was
shown between acupuncture and a penetrating sham method in
1 study (n = 18, WMD, 12.95; 95% CI of 34.05 to 8.15, P = 0.23)
(Fig. 3A3).29 Sensitivity analysis showed same results with above
meta-analysis regardless of study quality (data are not shown).
Two RCTs were not pooled (though they used pain intensity
by VAS as an outcome measure) because of insufficient
original data (no standard deviation),31 or dichotomous data
instead of continuous data.19
3.5.2. NRS for facial pain
Two RCTs20,28 reported on facial pain using NRS as pain
measurement. Although they reported a more favourable
effect of needle acupuncture in the original article, our
recalculation showed no significant difference between real
and sham groups (P = 0.30,20 P = 0.8428).
3.5.3. Muscle tenderness
Three RCTs21,30,31 assessed the effect of manual acupuncture
on muscle tenderness (4 point scale). All trials reported
favourable effects of acupuncture. One of them could not be
pooled in our meta-analysis, because of the scanty of the
original data (no standard deviation).31 A meta-analysis of
these data also showed significant, favourable effects of
needle acupuncture (2 studies, n = 46, SMD, 1.08; 95% CI of
1.88 to 0.28, P = 0.008, heterogeneity: x2 = 1.58, P = 0.21,
I2 = 37%) (Fig. 3B).
3.5.4. Mouth opening
Two RCTs30,31 reported the effect of manual acupuncture on
maximum MO measured in millimetres. Although one RCT
showed a favourable effect in the needle acupuncture group,
 journal of dentistry 39 (2011) 341–350 347
[()TD$FIG]

Fig. 3 – Meta-analysis of acupuncture for TMD. VAS: visual analogue scale.

348 journal of dentistry 39 (2011) 341–350
this study did not provide a clear standard deviation.31
Therefore, no further pooled analysis was completed due to
insufficient original data. Another RCT showed no statistical
difference between real and sham groups.30
4. Discussion
A total of 7 RCTs tested the effects of acupuncture compared to
sham acupuncture.19–21,28–31 This systematic review produced
limited evidence that acupuncture is more effective than
sham acupuncture in alleviating pain and masseter muscle
tenderness in TMD.
The methodological quality based on the Cochrane risk of
bias is generally moderate. Although each RCT adopted
assessor- and patient-blinding procedures, only four RCTs
used an appropriate sequence generation method.21,28,29,31
Only two RCTs employed allocation concealment.30,31 Trials
with inadequate sequence generation cause selection bias.32
Only two RCTs reported details of dropouts and withdra-
wals.20,31 This may lead to exclusion or attrition bias.32
All included trials suffered from lack of sufficient sample
size without power analysis, therefore, they are susceptible to
type II error.33 Small trials may also overestimate treatment
effects by about 30%.34 In our re-calculation of sample size
based on the mean difference value, at least 26 samples (in
each group) are needed for sufficient power (set at 80% power
and 0.05 in alpha error).35 None were met this power. Clinically
significant changes of pain VAS in patients with muscular
TMD was suggested by 24.2 mm,36 also in general pain severity
by 13–30 mm.37,38 Of included studies, the pain changes were
greater than 24.2 mm in two of included studies (one was
30.2 mm29 and the other was 33.2 mm31), however, four of
them were smaller.20,21,28,30 Thus, the reliability of the
evidence presented here is clearly limited.
In this review, various sham control methods have been
presented. The methods range from the use of a penetrating
needle on non-acupoints29 to non-penetrating needles on
acupoints19,21,30,31 or non-acupoints.20,28 ‘Real’ acupuncture
was found to be superior to non-penetrating sham acupuncture
control in pain outcomes20,21,28,30 but showed no difference from
penetrating sham acupuncture29 from our subgroup meta-
analysis. There are several possible explanations for these
findings; acupuncture might indeed be ineffective, alternatively,
penetrating sham control may have a similar effect to real one,
then the penetrating sham control may be invalid. Recently, one
study reported that a penetrating placebo needle could be
considered as having a similar effect of acupuncture, because it is
not inert.39
Furthermore, in another recent study,40 which was
reporting about relevance of non-penetrating sham acupunc-
ture, there was no significant difference in the correct
guessing of the real and sham acupuncture type amongst
blinded subjects. This may serve non-penetrating sham
acupuncture as a credible sham control. However, for
confirming this, rigorous studies between 2 types of sham
are needed for overcoming the controversy.
The double blind (i.e., blinding of the patient and the
evaluator) method was adopted in all of the trials. One study
suggested that either a penetrating or non-penetrating needle
placebo is effective in masking patients, whether the patients
were informed or not.41 Therefore, these sham control
methods seem to have potential uses in double blind studies.
However, the success of blinding was not confirmed in the
included studies. Future trials should consider testing and
reporting the success of blinding. Blinding of the therapist
remains an unresolved problem in studies of acupuncture.
Recent studies show that the ‘‘style’’ of the acupuncturists
significantly impacts on the clinical outcome.42 It would
therefore be important for future trials to find ways for
controlling for such confounding variables.
The optimal duration and number of acupuncture treat-
ments is an important clinical consideration.43 Four20,21,28,29 of
the seven studies used only single acupuncture treatments to
test acute effects, whilst the other three19,30,31 studies adopted
6–12 acupuncture treatments. Regardless of the number of
treatments, our meta-analysis of the pooled data showed
superior effects of acupuncture. These results either indicate
that acupuncture has a significant acute effect on TMD pain or
that the (non-blinded) acupuncturist’s ‘‘style’’ generates
immediate effects (see above).
Needle sensation (de-qi) was considered in one RCT,30 whilst
the other 6 trials did not report such details.19–21,28,29,31 None of
the studies reported the stimulation and manipulation meth-
ods. Recently, one f-MRI study reported that needle sensation
changed the brain response.44 Therefore, this seems to affect
treatment outcomes. However, there was no significant
difference between groups whether the trial adopted de-qi or
not. Hence, we do not know yet whether de-qi exerted an
important influence on the clinical outcomes or not.
A possible mechanism of acupuncture-mediated effects on
TMD, based on an electromyographic (EMG) study, is that
acupuncture causes the spinal cord and brain to release calming
agents such as serotonin, endorphins and neurotransmitters
with anti-inflammatory action.45 In addition, after treatment
with acupuncture, a better distribution of EMG activity was
observed that indicated a greater muscular balance with the
predominance of masseter muscles.45 However, none of these
effects have been independently confirmed.
Only one RCT addressed adverse events, but provided no
details.30 The fact that most studies fail to mention adverse
effects is remarkable. Adverse effects of acupuncture are well
known and thus need to be considered in any clinical research.
The lack of reporting of adverse effects seems to reflect the
often poor reporting of acupuncture trials. None of the
included trials reported ethical approval. This suggests the
need for an ethical review of acupuncture research.
In conclusion, the evidence for acupuncture for TMD
management, especially for TMJ and masseter muscles pain, is
weak. Thus, this conclusion is clearly limited due to the small
sample sizes and the small number of included RCTs.
Therefore, large-scale, rigorous studies with standardized
treatment method in high quality are needed to establish
whether acupuncture has definite therapeutic value.
Acknowledgments
Byung-Cheul Shin was supported for two years by a Pusan
National University Research Grant.
 journal of dentistry 39 (2011) 341–350
Appendix A. Supplementary data
Supplementary data associated with this article can be
found, in the online version, at doi:10.1016/j.jdent.2011.02.006.
Conflict of interest
The authors declare no conflicts of interest.
references
1. Scrivani SJ, Keith DA, Kaban LB. Temporomandibular
disorders. The New England Journal of Medicine 2008;359:
2693–705.
2. Leeuw R. American academy of orofacial pain: guidelines for
assessment, diagnosis and management. 4th ed. Chicago:
Quintessence Publishing; 2008. pp.129–75.
3. Lipton JA, Ship JA, Larach-Robinson D. Estimated prevalence
and distribution of reported orofacial pain in the United
States. The Journal of the American Dental Association
1993;124:115–21.
4. List T, Wahlund K, Wenneberg B, Dworkin SF. TMD in
children and adolescents: prevalence of pain, gender
differences, and perceived treatment need. Journal of
Orofacial Pain 1999;13:9–20.
5. Manfredini D, Piccotti F, Ferronato G, Guarda-Nardini L. Age
peaks of different RDC/TMD diagnoses in a patient
population. Journal of Dentistry 2010;38:392–9.
6. Jerjes W, Upile T, Abbas S, Kafas P, Vourvachis M, Rob J, et al.
Muscle disorders and dentition-related aspects in
temporomandibular disorders: controversies in the most
commonly used treatment modalities. International Archives
of Medicine 2008;1:23.
7. Syrop SB. Initial management of temporomandibular
disorders. Dentistry Today 2002;21:52–7.
8. DeBar LL, Vuckovic N, Schneider J, Ritenbaugh C. Use of
complementary and alternative medicine for
temporomandibular disorders. Journal of Orofacial Pain
2003;17:224–36.
9. Lin JG, Chen WL. Acupuncture analgesia: a review of its
mechanisms of actions. The American Journal of Chinese
Medicine 2008;36:635–45.
10. Zhao ZQ. Neural mechanism underlying acupuncture
analgesia. Progress in Neurobiology 2008;85:355–75.
11. Clinical research into the effectiveness of
acupuncture.White AR, Cummings M, Fishie J, editors. An
introduction to western medical acupuncture. Edinburgh, UK:
Churchill Livingstone/Elsevier; 2008. p. 93–118.
12. Linde K, Niemann K, Meissner K. Are sham acupuncture
interventions more effective than (other) placebos? A re-
analysis of data from the Cochrane review on placebo
effects. Research in Complementary Medicine 2010;17:259–64.
13. Sites BD, Neal JM. Placebo or intervention? Is it all a sham?
Anaesthesia 2011;66:73–5.
14. Greene CS, Goddard G, Macaluso GM, Mauro G. Topical
review: placebo responses and therapeutic responses. How
are they related? Journal of Orofacial Pain 2009;23:93–107.
15. Cho SH, Whang WW. Acupuncture for temporomandibular
disorders: a systematic review. Journal of Orofacial Pain
2010;24:152–62.
16. Ernst E, White AR. Acupuncture as a treatment for
temporomandibular joint dysfunction: a systematic review
of randomized trials. Archives of Otolaryngology—Head & Neck
Surgery 1999;125:269–72.
349
17. La Touche R, Angulo-Diaz-Parreno S, de-la-Hoz JL,
Fernandez-Carnero J, Ge HY, Linares MT, et al. Effectiveness
of acupuncture in the treatment of temporomandibular
disorders of muscular origin: a systematic review of the last
decade. Journal of Alternative and Complementary Medicine
2009;16:107–12.
18. La Touche R, Goddard G, De-la-Hoz JL, Wang K, Paris-
Alemany A, Angulo-Diaz-Parreno S, et al. Acupuncture in
the treatment of pain in temporomandibular disorders: a
systematic review and meta-analysis of randomized
controlled trials. The Clinical Journal of Pain 2010;26:
541–50.
19. Katsoulis J, Ausfeld-Hafter B, Windecker-Getaz I, Katsoulis
K, Blagojevic N, Mericske-Stern R. Laser acupuncture for
myofascial pain of the masticatory muscles. A controlled
pilot study. Schweiz Monatsschrift fuer Zahnmedizin
2010;120:213–25.
20. Shen YF, Goddard G. The short-term effects of acupuncture
on myofascial pain patients after clenching. Pain Practice
2007;7:256–64.
21. Simma I, Gleditsch JM, Simma L, Piehslinger E. Immediate
effects of microsystem acupuncture in patients with
oromyofacial pain and craniomandibular disorders (CMD): a
double-blind, placebo-controlled trial. British Dental Journal
2009;207:E26.
22. Liberati A, Altman DG, Tetzlaff J, Mulrow C, Gotzsche PC,
Ioannidis JP, et al. The PRISMA statement for reporting
systematic reviews and meta-analyses of studies that
evaluate health care interventions: explanation and
elaboration. PLoS Medicine 2009;6:e1000100.
23. Liu J, Fei Y, Alraek T. Acupuncture for treatment of erectile
dysfunction (protocol). Cochrane Database of Systematic Review
2008:CD007241.
24. Higgins J, Green S, Higgins JPT, Green S, editors. Cochrane
handbook for systematic reviews of interventions version 5.0.2
[updated September 2009]. The Cochrane Collaboration;
2009 www.cochrane-handbook.org.
25. Follmann D, Elliott P, Suh I, Cutler J. Variance imputation for
overviews of clinical trials with continuous response. Journal
of Clinical Epidemiology 1992;45:769–73.
26. Higgins J, Thompson S, Deeks J, Altman D. Statistical
heterogeneity in systematic reviews of clinical trials: a
critical appraisal of guidelines and practice. Journal of Health
Services Research & Policy 2002;7:51–61.
27. Elsharkawy TM, Ali NM. Evaluation of acupuncture and
occlusal splint therapy in the treatment of
temporomandibular joint disorders. Egyptian dental journal
1995;41:1227–32.
28. Shen YF, Younger J, Goddard G, Mackey S. Randomized
clinical trial of acupuncture for myofascial pain of the jaw
muscles. Journal of Orofacial Pain 2009;23:353–9.
29. Goddard G, Karibe H, McNeill C, Villafuerte E. Acupuncture
and sham acupuncture reduce muscle pain in myofascial
pain patients. Journal of Orofacial Pain 2002;16:71–6.
30. Schmid-Schwap M, Simma-Kletschka I, Stockner A,
Sengstbratl M, Gleditsch J, Kundi M, et al. Oral acupuncture
in the therapy of craniomandibular dysfunction syndrome –
a randomized controlled trial (RCT). The Middle-European
Journal of Medicine 2006;118:36–42.
31. Smith P, Mosscrop D, Davies S, Sloan P, Al-Ani Z. The
efficacy of acupuncture in the treatment of
temporomandibular joint myofascial pain: a randomised
controlled trial. Journal of Dentistry 2007;35:259–67.
32. Juni P, Altman DG, Egger M. Systematic reviews in health
care: assessing the quality of controlled clinical trials. British
Medical Journal 2001;323:42–6.
33. Litiere S, Alonso A, Molenberghs G. Type I and Type II error
under random-effects misspecification in generalized linear
mixed models. Biometrics 2007;63:1038–44.
350 journal of dentistry 39 (2011) 341–350
34. Moore A, Edwards J, Barden J, Mcquay H. Bandolier’s little
book of pain. Oxford: Oxford University Publishing; 2003.
35. Chow SC, Shao J, Wang H. Sample size calculations in
clinical research. 2nd ed. Boca Raton, FL: Chapman & Hall/
CRC; 2008. pp. 49–81.
36. van Grootel RJ, van der Glas HW. Statistically and clinically
important change of pain scores in patients with
myogenous temporomandibular disorders. European Journal
of Pain 2009;13:506–10.
37. Gallagher EJ, Liebman M, Bijur PE. Prospective validation of
clinically important changes in pain severity measured on a
visual analog scale. Annals of Emergency Medicine
2001;38:633–8.
38. Lee JS, Hobden E, Stiell IG, Wells GA. Clinically important
change in the visual analog scale after adequate
pain control. Academic Emergency Medicine 2003;10:
1128–30.
39. Lund I, Lundeberg T. Are minimal, superficial or sham
acupuncture procedures acceptable as inert placebo
controls? Acupuncture Medicine 2006;24:13–5.
40. Lee H, Bang H, Kim Y, Park J, Lee S, Park HJ. Non-penetrating
sham needle is it an adequate sham control in acupuncture
research? Complementary therapies in medicine 2011;19:S41–8.
41. Takakura N, Yajima H. A placebo acupuncture needle with
potential for double blinding – a validation study.
Acupuncture in Medicine 2008;26:224–30.
42. Witt CM, Ludtke R, Wegscheider K, Willich SN. Physician
characteristics and variation in treatment outcomes: are
better qualified and experienced physicians more
successful in treating patients with chronic pain with
acupuncture? Journal of Pain 2010;11:431–5.
43. Slopeck A, Feng HT. Qi, time and acupuncture. Journal of
Aupuncture and Tuina Science 2010;8:199–264.
44. Asghar AU, Green G, Lythgoe MF, Lewith G, MacPherson H.
Acupuncture needling sensation: the neural correlates of
deqi using fMRI. Brain Research 2009;1315:111–8.
45. Rancan SV, Bataglion C, Bataglion SA, Bechara OM, Semprini
M, Siessere S, et al. Acupuncture and temporomandibular
disorders: a 3-month follow-up EMG study. Journal of
Alternative and Complementary Medicine 2009;15:1307–10.